One scalable reward system: our expertise

    Our favorite airline sells tickets through a network of its own and independent agents and partners, cooperates with all market participants: agents, travel portals, tour operators. Clients are both individual travelers and corporate clients. To stimulate sales, the airline has several reward systems and loyalty programs for agents and customers. They appeared at different times, initiated by different departments of the airline and accounting for each program is independent.

    As a result, to get a consolidated picture of the costs and effectiveness of all loyalty systems was quite problematic, because this was done manually.

    The introduction of common metrics and accounting rules allowed us not only to automate this process and get analytics in real time, but also to plan the development of individual loyalty tools taking into account the effectiveness of the system as a whole.

    We developed a remuneration conditions constructor, essentially a billing system that allows you to configure any condition for each program and calculate the general economy of loyalty systems based on historical data on sales (or services provided) for past periods.

    Condition constructor

    The product is an algorithm for calculating rewards and a web service (website) on which company managers can assign rewards to partners or customers and calculate the economy of their decision on historical data.


    To create a constructor of conditions, we made an “absolute catalog” of the characteristics of the service - in fact, we decomposed the reward system “into atoms”.

    • We have identified the entire list of characteristics of the sale (service): roughly speaking, what type of transportation we sell, which agent sells, we sell offline or online, we transport ourselves or we sell the transportation of another airline, etc. etc.
    • Identified accrual strategies: we charge increased or decreased interest, bonuses, or money.
    • We determined how the rules can work: include a condition, exclude a condition, set a priority.

    And now from these "atoms" we can configure any reward.

    Web service

    The specialist’s workplace now is an automated portal.

    • On the “Template Designer” page, a specialist formulates a rule for calculating rewards. The correct form of the rule is already in the template. And the options for filling in the variable data are the same. The specialist simply selects from the list. At the final stage, the service verifies the parameters that the specialist sets for completeness and consistency, and saves.image
    • In the “Remuneration conditions constructor” section, the specialist assigns (applies) the finished template to the partner or client, or to the whole group. image
    • The miscalculation functional of the model
      Before applying the condition, the specialist can calculate the estimated economy of the conditions that he sets on the archived data.
    • Application of the condition The condition
      designer is directly linked to the airline's accounting system. In online mode, it reports data on current conditions for agents. And on the basis of this information, the accounting system issues invoices to agents (or accrues bonuses in case of loyalty systems).
    • Adding a new loyalty program
      Now it doesn’t matter how many loyalty programs exist in parallel in the company and whether new ones appear. Since our service (sale) is already decomposed into “atoms”, each new loyalty program simply uses a relevant set of characteristics and applies the necessary strategies for calculating rewards. And the model continues to consider the overall economy of all loyalty systems.

    So our solution helped bring together all reward systems and loyalty programs and unified all settlement services, and miscalculation of the model allows you to compare the cost of different conditions and choose the most significant for the airline.

    Also popular now: